AI Will Kill Medium, but There Are Viable Solutions – Part 1:

Summary of a 16 hour interview with Dr Mehmet Yildiz in six part

and

Jan 29, 2025

We summarized the interview in an interactive and educational audio using Google’s KLM technology in 15 minutes which you can listen before or after reading the questions in the first part of of this strategic leadership interview series.

1×

0:00

-15:11

Why Medium Is in Trouble—and How It Can Still Be Saved

As senior editors of ILLUMINATION and founding members of the Substack Mastery Boost Pilot, we have observed an undeniable truth: AI is neither the enemy nor the savior—it is merely a tool. The real question is not whether AI will overrun Medium, but whether Medium can adapt before it is too late. Right now, the platform is in deep trouble, and its future hinges not on AI itself, but on its own response to it.

The recent drama surrounding payments and AI-driven content floods has left writers frustrated, publications overwhelmed, and readers disillusioned. To address these urgent concerns, we conducted an intensive 16-hour interview with our chief editor,

Dr Mehmet Yildiz

, after his work hours via Slack Huddle. This was not just another conversation—it was a deep-dive leadership session, analyzing Medium’s flaws, opportunities, and the roadmap needed to prevent its decline.

We specifically chose Dr. Yildiz for this leadership discussion because of his scholarly, scientific, and technological expertise, combined with content strategy, writing, and marketing skills rooted in an ethnographic approach. Having been on Medium for years, he has witnessed its evolution firsthand, backing his observations with empirical research and data produced by his data science team—including experts from MIT who support his vision.

Why We Created This Interview Series

Our team has observed an alarming rise in AI-generated drafts submitted to our publications. Every single day, we review and return over 100 AI-generated pieces, many of which have been humanized using deceptive techniques to bypass detection.

If one publication alone is drowning in this volume of AI-generated content—and writers typically submit their best pieces to publications—then the true scale of AI-generated material flooding Medium is staggering. Multiply this by 20,000+ publications and add the self-published AI-generated flood, and what emerges is a relentless tsunami of machine-written, engagement-optimized filler with zero human insight.

Worse, these AI-generated pieces are not just slipping through unnoticed—they are actively deceiving Medium’s detection systems. Tools like Copyleaks, which Medium appears to trust, show a 0% AI detection rate on manipulated AI content when we tested them. The reason? AI-generated articles are being disguised by “humanizers”—cheap, deceptive tools that rewrite AI content just enough to evade detection while contributing nothing of real value.

This is not just an influx of low-quality content—it is a systemic breakdown. A platform that was once a haven for authentic writing is now drowning in algorithmically-generated noise, and unless decisive action is taken, Medium risks becoming indistinguishable from AI content farms, its credibility irreparably damaged.

But our human editors can spot this AI-generated nonsense instantly—no advanced detection tools required. The irony? Medium does not even have as many dedicated volunteer editors as ILLUMINATION publications. If we, as an independent community, can detect this deception so easily, why is a multi-million-dollar platform struggling to do the same?

At this rate, Medium is spiraling toward becoming an unregulated AI content mill, where machine-generated filler saturates the platform, drowning out real voices. If nothing changes, its credibility will collapse, and the exodus of serious writers and readers will only accelerate. Once trust is lost, there is no undo button.

But Dr. Yildiz urges us and Medium leadership team to rise above AI-phobia. Fear and reactionary policies will not save Medium—only strategic, transparent, and human-led solutions can. Instead of fighting AI blindly, he proposes a thoughtful, structured approach—one that regulates AI’s role rather than letting it hijack the platform:

✔ A sophisticated AI detection system—not just a flawed, algorithmic filter, but a technology-driven, human-assisted approach that accurately identifies AI-generated content and redirects it appropriately without punishing genuine writers.

✔ Human oversight as a safeguard—because AI detection tools are not infallible. Skilled human editors must intervene where algorithms fail, preventing misclassification and ensuring that real writers are not unfairly penalized.

✔ Full transparency in AI integration—so that readers are not deceived into consuming AI-generated content disguised as human work. Medium must empower its audience with clear content distinctions, allowing them to make informed choices about what they engage with.

This is not just a moment of crisis but is a defining turning point for Medium and everyone invested in its future.

If Medium’s leadership continues to ignore the growing concerns of its most dedicated writers, editors, thought leaders, and discerning readers, the platform will not survive. It will collapse—not because of AI, but because of its own refusal to adapt responsibly.

But there is still time and hope. If Medium listens, acts decisively, and evolves, it has a rare opportunity to lead the ethical integration of AI in digital publishing. The choice is clear: innovate with transparency and integrity, or watch the platform fade into irrelevance.

A Rare Opportunity to Learn from a Thought Leader

Today, we are giving our free subscribers an exclusive preview (the first part) of this powerful interview series. However, given the significant time investment from our editorial team—including 16 hours of Dr. Yildiz’s expertise, strategy, and insight—the full series will be available as premium content for our paid members.

This is not an ordinary interview. It is a well-thought-out leadership workshop—an invaluable resource for writers, editors, and discerning readers who want to stay ahead in the AI-driven content era. More importantly, it is a wake-up call for Medium’s leadership:

✔ The time for arrogance is over. It is time for humility and transparency.
✔ The time for real, data-driven solutions is now.

If Medium’s leaders truly care about the platform’s future, they would pay attention to thought leaders like Dr. Yildiz, rather than dismiss the voices trying to help.

So sit back, absorb these critical insights, and be part of the solution. The future of digital publishing is being written right now—and it is up to us to shape it with intelligence, integrity, and a commitment to human creativity.

Part 1: The Core Debate

Question 1: Dr Yildiz, before starting this important part, can you tell us why you stopped writing on Medium but why do you still support Medium?

I made the painful decision to stop writing on Medium after their latest algorithmic shift, which used "spam and scam" as a pretext to slash the earnings of genuine writers. This is not just about money—it is about the blatant disregard for the very people who built this platform’s success. Until Medium corrects this, I refuse to contribute further.

Yet, despite my disappointment, I cannot abandon this platform. Six years ago, I was drawn to Medium’s vision and poured my heart and soul into it. To watch it crumble due to poor leadership is heartbreaking. The thought of this vibrant community losing such a powerful opportunity because of mismanagement is unbearable.

Medium has immense potential—innovative features, a dedicated user base, and a legacy of quality writing. But its leaders squandered these strengths, clinging to arrogance rather than listening to the brilliant minds who genuinely cared. Instead of leveraging their most passionate members, they dismissed invaluable feedback and alienated the very people who could have made Medium unstoppable.

If I were in charge, I would have transformed this small startup into a billion-dollar unicorn. And if Medium collapses under the weight of its own missteps, I will not let its essence die—I will create something better.

But my hope remains that Medium can still be saved. That is why, despite the platform’s poor treatment of me, my publications, my volunteer editors, and our writing community, I continue to support this incredible network of creators. I want it to thrive—not for the sake of its leadership, but for the writers, readers, and dreamers who make it truly special.

Question 2: Is AI a real threat to Medium?

Absolutely—but not in the way people might think. AI itself is not a villain lurking in the shadows, waiting to destroy Medium. The real danger is in how Medium responds to it. If the platform passively allows an unchecked flood of AI-generated content, it will suffocate human creativity.

Readers will struggle to find authentic voices amid a sea of algorithmically optimized articles designed purely for engagement and money making by oportunists or scammers who are currently taking the money from the payment pool depriving authentic writers. Trust will erode, discovery will collapse, and eventually, both writers and audiences will drift away.

But AI is not inherently destructive—it is a tool. If Medium uses it wisely, AI could actually enhance the platform by improving discovery, reducing tedious tasks for writers, and helping creators refine their work. The challenge is balance. Right now, AI-generated content is poised to overrun the platform, and if Medium does not set clear boundaries, human storytelling—the very essence of what makes Medium valuable—will be lost.

Question 3: Is there a solution to it at a high level?

Yes, a smart solution is to structure the platform into two distinct but interconnected spaces: one exclusively for human-authored stories and another for AI-generated content. I will articulate it in next parts with details.

This strategic approach would preserve the integrity of human storytelling while still allowing AI-driven content to exist transparently.

Medium has an opportunity to lead the ethical integration of AI in publishing, but if it does nothing, it risks turning into yet another content farm—indistinguishable from the algorithmic noise AI is so good at generating.

Question 4: What is the worst-case scenario for Medium?

If AI-generated content floods the platform unchecked, authentic engagement will dwindle, and genuine writers will leave. This is not just a theoretical risk—it is already happening. Hundreds of talented writers have walked away, and while the full impact is not immediately visible, readers will soon feel the void. The absence of unique, thoughtful voices will erode trust, making Medium indistinguishable from AI-driven content mills.

Losing credibility is not a slow decline—it is an irreversible tipping point. If Medium does not act swiftly, it could collapse as a meaningful space for both writers and paying readers. AI-generated content, optimized for clicks rather than depth, will dominate the feed, turning Medium into a repetitive echo chamber where engagement numbers might look strong, but actual value diminishes.

The solution is not to resist AI but to regulate it. Medium must acknowledge the reality and implement a structured system that separates human and AI content. A dynamic AI detection and sorting mechanism—one that evolves with user feedback—could prevent AI from unfairly outcompeting human creativity. Otherwise, Medium risks becoming a cautionary tale of how an innovative platform, once built on original storytelling, was undone by its failure to adapt to the very technology shaping the future of publishing. I will expand on this further in the upcoming questions.

Question 5: Does AI expose weaknesses in Medium’s platform?

Yes, but not in the way some might assume. AI did not create Medium’s problems—it simply magnified them. The platform has long been overly reliant on vanity metrics like claps and scattered highlights, which prioritize virality over substance.

AI-generated content is now exploiting these flaws more efficiently than human writers ever could. It is optimized for algorithms, not for depth, and it easily tricks Medium’s system into rewarding engagement-driven, surface-level writing over originality and insight. Besides claps, highlights and artificial coments are sold in Fiverr and scammers use them to steal earning of authentic writers. This is so sad and heartbreaking. I escalated this to the leadership of the platform but they did not even acknowledged let alone thank me and my team who removed sellers of artifical metrics on my Slack workspace and social media.

If Medium wants to remain a meaningful space for writers and readers, it must rethink its approach. Instead of claps or highlights dictating virality and earnings, the platform should emphasize authenticity with reading times of paid members for monetization of efforts, well-researched perspectives, and true audience engagement. One way to achieve this is by creating distinct spaces for human-written and AI-generated content, ensuring that AI does not outcompete human creativity in discovery and visibility.

However, separating AI from human content is not enough—it requires a robust, continuously improving verification system. AI detection tools today are inconsistent, misidentifying polished human writing as machine-generated.

A smarter approach would blend AI-assisted classification with human oversight, refining accuracy over time. Without such measures, Medium risks becoming yet another AI-content mill—drifting further from the platform that once championed thought-provoking, human-driven storytelling.

Question 6: Could Medium turn into an AI-driven content farm?

Without proper regulation, Medium is on track to becoming yet another AI-powered content mill—just like the SEO-driven platforms that thrived on quantity over quality before collapsing under their own weight. Readers are discerning; once they realize the platform is flooded with AI-generated, engagement-optimized content, trust will erode. And when trust disappears, so does meaningful engagement.

AI-generated articles might drive short-term traffic, but they lack the depth, nuance, and originality that keep readers coming back. If human writers continue to leave in frustration, Medium risks becoming a hollow, algorithm-fed machine that churns out content without substance. The decline would be slow at first, but once credibility is lost, rebuilding it will be nearly impossible.

The viable solution is to acknowledge the inevitability of AI while ensuring it does not dilute human creativity. A dedicated AI-content section would allow transparency while preventing AI from unfairly dominating discovery and engagement. Medium still has a chance to lead in ethical AI integration—if it acts decisively. I believe this sums up the core issue. Now, let’s explore the details in part II.

Thank you for reading the first part of this interview.

We will also compile this interview in a digital and audio books and will make it available to our paid members. It will be a $100 value which is half of yearly subscription of our ILLUMINATION Library for Freelance writers which will provide many more premium content and Tier 2 level benefits of the Substack Mastery program.

Call to Action

Please take a moment to participate in the poll Dr. Yildiz posted yesterday. We were genuinely surprised to see that 20% of writers feel comfortable having their content assessed by Copyleaks. This raises an important question:

Did these writers have a chance to carefully consider the three key points Dr. Yildiz outlined? If not, we encourage you to review them again and reflect on their implications.

If you chose YES, we would truly appreciate hearing your perspective. This is not about changing your mind—it is about understanding your rationale so we can learn from your insights. Thoughtful discussions help all of us navigate these challenges more effectively, and your input is valuable.

So, whether you agree or disagree, please share your thoughts in the comments. Your voice matters, and we are here to listen. Here is the link to the newsletter with a poll: